Abstract
The idea of enriching a cultural heritage visitor's experience by employing mobile devices such as smartphones and tablets is an active research domain. This work presents the experience and feedback gained by one year of operation of an experimental information system that aims in enhancing the experience of a visitor on his/her demand. The evaluated system is based on the use case-scenario where a mobile device is used to capture and decode passive markers (such as QR-codes) that are located on facades of selected important buildings in the old town of Xanthi, Greece, in order to enrich the visitor's experience through a broad context of historical and architectural information delivered over the Web.
Introduction: The Challenge of Outdoor Cultural Enrichment
Enhancing the "visitor experience" at cultural heritage sites often involves a difficult trade-off between immersion and infrastructure. Traditional solutions—like proprietary audio guides or high-tech augmented reality (AR) rigs—are expensive to maintain, prone to hardware failure, and can be cumbersome for the user.
This research addresses the challenge of creating a scalable, low-cost information system that works in uncontrolled outdoor environments. The authors focus on an "on-demand" model: instead of pushing information to a user based on GPS (which can be battery-intensive and intrusive), the system allows visitors to "pull" information only when a specific landmark captures their interest.
Methodology: Implementing Passive Markers
The researchers implemented a system based on passive markers—specifically QR codes. Unlike "active" markers (like Bluetooth beacons), passive markers require no power source and are extremely inexpensive to produce.
Key Technical Components: * Physical Infrastructure: 13 Points of Interest (POIs) were marked with A4-sized metal plates. These were covered with an antiglare film to ensure the QR codes remained readable under varying outdoor lighting conditions. * Backend Architecture: A standard LAMP-style stack (PHP, MySQL) served content. The system used URI-based queries to identify the specific POI and the user's preferred language. * User Interaction: The workflow was designed to be "simplistic": Scan QR -> Decode -> URI Request via Web Browser -> Content Delivery. This avoided the need for users to download a dedicated, site-specific app.
Study Area and Data Collection
The "dataset" for this study consists of quantitative usage logs collected over one full year of operation in the old town of Xanthi, Greece. The scope included 13 distinct architectural POIs, ranging from churches and mansions to public squares.
Over the 12-month period, 879 unique content requests were logged and analyzed to determine patterns in visitor behavior, language preference, and device usage.
Key Insights and Results
- The "Crossroad" Effect: Usage wasn't uniform. POIs located on crossroads or in highly photographed areas saw significantly higher engagement than those in narrow side streets. This suggests that physical placement and "visual prominence" are as important as the digital content itself.
- Vandalism as a Design Factor: Within the first month, 15% of the markers were damaged or removed. For unattended outdoor systems, physical durability and a plan for replacement are critical, non-technical requirements.
- The "Pull" Preference: The steady stream of hits throughout the year—including late-night hours—suggests that visitors appreciate the agency of choosing when to interact with the system.
- Hardware Agnosticism: The logs showed a massive fragmentation of Android versions (from 2.2.1 to 4.1.2). By using a web-based delivery method instead of a native app, the system remained accessible to users with older hardware.
Quantitative Results: * Engagement: 879 unique hits over 12 months. * Demographics: 74% of requests were for Greek content, while 26% were for English. * Platform Dominance: Android accounted for 63% of users, followed by iOS at 25%. * Performance: On a standard 3G network using a mid-range smartphone, the average time from scan to content display was approximately 8 seconds.
Limitations and Future Work
The authors identify several hurdles for future iterations:
- Connectivity Barriers: While the town has free Wi-Fi, the registration process can be a deterrent. Conversely, using mobile data can lead to high roaming charges for international tourists.
- Content Depth: The current pilot primarily delivered text and images. Future versions should incorporate more "multilateral" content, such as semi-automated guided tours.
- Marker Aesthetics: While the A4 plates were necessary for visibility, there is a need to balance "being obvious" with the aesthetic preservation of historical facades. Future work may look at smaller or more discreet markers integrated into existing signage.
Summary updated